Plasma preparation from whole blood using ultrasound.

A technique to efficiently separate plasma from human whole blood is described. Essentially, 3-mL samples are held on the axis of a tubular transducer and exposed for 5.7 min to an ultrasonic standing wave. The cells concentrate into clumps at radial separations of half wavelength. The clumps grow in size and sediment under gravity. A distinct plasma/cell interface forms as the cells sediment. The volume of clarified plasma increases with time. The separation efficiencies of transducers of 29-mm and 23-mm internal diameters driven, by test equipment, at radial resonances of 3.4 and 1.5 MHz, respectively, were compared. The average efficiency of separation was 99.6% at 1.5 MHz and 99.4% with the 3.4-MHz system. The cleared plasma constituted 30% of the sample volume at 1.5 MHz and 25% at 3. 4 MHz. There was no measurable release of haemoglobin or potassium into the suspending phase, indicating that there was no mechanical damage to cells at either frequency. A total of 114 samples from volunteers and patients were subsequently clarified in a 1.5-MHz system driven by an integrated generator. The average efficiency of clarification of blood was 99.76% for the latter samples. The clarification achieved is a significant improvement on that previously reported (98.5%) for whole blood exposed to a planar ultrasonic standing wave field (Peterson et al. 1986). We have, therefore, now achieved a six-fold reduction of cells in plasma compared to previous results.

Clarification of small volume microbial suspensions in an ultrasonic standing wave.

The removal of Saccharomyces cerevisiae and Escherichia coli from 2.5 ml suspensions in ultrasonic standing wave formed at 1 or 3 MHz has been characterized. The standing wave was set up by a plane transducer and reflector mounted in the vertical plane. Cells in the ultrasonic field first concentrated in vertical planes at half wavelength separations. The ultrasound was then pulsed to allow clumps of concentrated cells to sediment in a controlled way during the short 'off' intervals. Yeast removal from suspension at a concentration of 3 x 10(9) ml-1 (14% volume v/v) was 99.5% in a total time of 4.5 min. Almost total (99.5%) clarification of prokaryote (E. coli) suspension was achieved here for the first time in a standing wave field. The clarification of a 1.3 x 10(11) ml-1 (16% v/v) E. coli suspension occurred over 11.5 min. The period decreased to 7 min in the presence of a polycationic flocculant, polyethyleneimine. The implications of the results for design of systems to further reduce clarification times are discussed. Removal efficiency for both S. cerevisiae and E. coli decreased with decrease in cell concentration. This concentration dependence is shown not to be simply a consequence of acoustic interaction between single cells. Flow cytometry of stained cells detected no loss of cell viability arising from the ultrasonic procedure.

Filtration of bacteria and yeast by ultrasound-enhanced sedimentation.

Continuous flow filtration of suspensions of eukaryotic cells by ultrasonic standing wave enhanced sedimentation has recently been reported. The filtration efficiency for Escherichia coli in such a filter has been characterized at frequencies of 1 and 3 MHz in the present work and compared with results for Saccharomyces cerevisiae. The yeast can be filtered at greater than 99% efficiency at a flow rate of 5 ml min-1 at either frequency. The filtration efficiency of the smaller E. coli at 3 MHz is in excess of 80% at concentrations in the region of 10(10) ml-1 but decreased at lower concentrations. However, E. coli in a mixed suspension with yeast were, because of inter-particle interactions, removed with the filtrate at an efficiency ranging from 80 to 50% over the eight orders of bacterial concentrations tested (down to 10(3) ml-1) at 3 MHz. Quantitative considerations show that poor filtration of pure suspensions of the smaller cells at the lower frequency arises because, at reasonable flow rates, the residence time is not sufficient for the cells to reach the pressure nodal cell concentration regions. The filtration efficiencies of both cell types are comparable at 3 MHz. It is suggested that the more comparable efficiencies arise because concentration regions are narrower at the high frequency and Stokes drag by the filter bulk flow inhibits sedimentation of the concentrated cells.